NeuralEnsemble
Error loading plx file
I'm trying to load a .plx file and I'm getting an error when using .read(). This error seems to be specific for this file. However, I know the file is not corrupted because I can open it with other software. I have tested a couple more .plx files and they seem to work fine. I'm happy to share the file.
Here's my code:
from neo.io import PlexonIO
file = "my_file_location.plx"
reader = PlexonIO(filename = file)
This runs fine, although it outputs this Warning:
C:\Users\myusername\.conda\envs\neo2\lib\site-packages\neo\rawio\plexonrawio.py:163: RuntimeWarning: divide by zero encountered in double_scalars
gain = global_header['SlowMaxMagnitudeMV'] / (
However, when I run reader.read() I get this error:
---------------------------------------------------------------------------
NameError Traceback (most recent call last)
File ~\.conda\envs\neo2\lib\site-packages\quantities\registry.py:20, in UnitRegistry.__Registry.__getitem__(self, string)
19 try:
---> 20 return eval(string, self.__context)
21 except NameError:
22 # could return self['UnitQuantity'](string)
File <string>:1, in <module>
NameError: name 'inf' is not defined
During handling of the above exception, another exception occurred:
LookupError Traceback (most recent call last)
Input In [52], in <cell line: 1>()
----> 1 reader2.read()
File ~\.conda\envs\neo2\lib\site-packages\neo\io\baseio.py:124, in BaseIO.read(self, lazy, **kargs)
121 if (hasattr(self, 'read_all_blocks') and
122 callable(getattr(self, 'read_all_blocks'))):
123 return self.read_all_blocks(lazy=lazy, **kargs)
--> 124 return [self.read_block(lazy=lazy, **kargs)]
125 elif Segment in self.readable_objects:
126 bl = Block(name='One segment only')
File ~\.conda\envs\neo2\lib\site-packages\neo\io\basefromrawio.py:166, in BaseFromRaw.read_block(self, block_index, lazy, create_group_across_segment, signal_group_mode, load_waveforms)
164 # Read all segments
165 for seg_index in range(self.segment_count(block_index)):
--> 166 seg = self.read_segment(block_index=block_index, seg_index=seg_index,
167 lazy=lazy, signal_group_mode=signal_group_mode,
168 load_waveforms=load_waveforms)
169 bl.segments.append(seg)
171 # create link between group (across segment) and data objects
File ~\.conda\envs\neo2\lib\site-packages\neo\io\basefromrawio.py:235, in BaseFromRaw.read_segment(self, block_index, seg_index, lazy, signal_group_mode, load_waveforms, time_slice, strict_slicing)
233 for sub_stream in sub_streams:
234 stream_index, inner_stream_channels, name = sub_stream
--> 235 anasig = AnalogSignalProxy(rawio=self, stream_index=stream_index,
236 inner_stream_channels=inner_stream_channels,
237 block_index=block_index, seg_index=seg_index)
238 anasig.name = name
240 if not lazy:
241 # ... and get the real AnalogSignal if not lazy
File ~\.conda\envs\neo2\lib\site-packages\neo\io\proxyobjects.py:147, in AnalogSignalProxy.__init__(self, rawio, stream_index, inner_stream_channels, block_index, seg_index)
145 str_units = ensure_signal_units(sig_chans['units'][0]).units.dimensionality.string
146 gain0 = sig_chans['gain'][0]
--> 147 self._raw_units = pq.CompoundUnit(f'{gain0}*{str_units}')
148 else:
149 self._raw_units = None
File ~\.conda\envs\neo2\lib\site-packages\quantities\unitquantity.py:376, in CompoundUnit.__new__(cls, name)
375 def __new__(cls, name):
--> 376 return UnitQuantity.__new__(cls, name, unit_registry[name])
File ~\.conda\envs\neo2\lib\site-packages\quantities\registry.py:60, in UnitRegistry.__getitem__(self, label)
57 if "%" in label: label = label.replace("%", "percent")
58 if label.lower() == "in": label = "inch"
---> 60 return self.__registry[label]
File ~\.conda\envs\neo2\lib\site-packages\quantities\registry.py:23, in UnitRegistry.__Registry.__getitem__(self, string)
20 return eval(string, self.__context)
21 except NameError:
22 # could return self['UnitQuantity'](string)
---> 23 raise LookupError(
24 'Unable to parse units: "%s"'%string
25 )
LookupError: Unable to parse units: "inf*dimensionless"
Hi @AlexLM96 It looks like your signal does not have gain attached, as neo is attempting to use an infinite gain. I am surprised this is not an issue in other softwares. How do these handle this case? Have you tried loading the data in non-lazy mode? Does this also fail? Would it be possible for you to share the respective file for debugging purposes?
It still fails in non-lazy mode. And sure, I'll email you the file. Thanks!
@AlexLM96 If the issue persists feel free to upload the file using the link provided above